This is Part I of three hypothesis I have. I welcome your thoughts.
Hypothesis I: Even within the same genotype, there are different deficiencies in the CFTR function. Some f508d mutations cannot fold properly, others cannot sequence the folding properly, others cannot gate the protein properly, and then there is some combination of all of this, at varying levels. Thus, whether those homozygous for f508d will need a corrector, a potentiator, or a “sequencer” will vary and there is no way currently to identify who has which defect/combination of defect. But this variation of defects within the f508d population make it more likely that the 809/661 Kalydeco combo will, overall, provide the necessary benefit to justify approval.
Scientific Studies & Hearsay:
1. At the last CF Conference it was made clear that researchers believe there are three defects in f508d CFTR: First, the protein does not fold properly; second the folding must be done in a proper order but it is not sequenced properly; and third the protein does not gate properly. Researches indicated that they believe there will need to be three compounds to address these defects.
2. Some doctors have explained to Cfers that for some double deltas, once the protein reaches the cell surface, it does not have an issue with gating, i.e., it can gate fine on its own. I found this explanation both the clearest and most entertaining:
For CFTR, imagine there is a quality control agent (for the younger ones in the crowd -- the bouncer outside the door at the hottest club in town...). Ppl with full functioning CFTR's are the hot chicks in the queue who get right in no problem, even though there is a line two hours long (quit bragging - you know who you are!). For those with both copies, the bouncer is not letting us in because we don't 'look' the part; we have money in our wallets and love the hip music, but our shirts aren't tucked in... we showed up in sneakers... they see we have Spandau Ballet on our iPODs, etc. There is no way into the club! 661 taps the bouncer on the shoulder, he looks the other way, so we can get in! some of us, there is a theory, have credit cards with no limits (think of this as the chloride channels working just fine...). If *some* of the folks get past the bouncer (the protein makes it to the top of the cell) and you have no credit on your credit cards (chloride channels not functioning as advertised) then 770 OR SOMETHING LIKE 770 has to quickly call the bank and extend your credit.For the record, it wasn't explained to me like a night of clubbing and dancing with no worries, but it kinda left my fingertips like that... maybe because i so desperately want this to work so we CAN ALL GO OUT CLUBBING, for real, WITH NO WORRIES, DANCING AND DANCING, with lots of breaths and of course, no Spandau Ballet...
3. I have also heard/read that some double deltas have some protein at the cell surface, which means that for some there is a least some folding/sequencing function and possibly some gating function.
4. Scientists also have said that genotype does not determine phenotype and actually the lowest correlation between genotype and phenotype is for lung function (pancreatic function has the highest correlation). This fact might explain why FEV improved so dramatically when SC didn’t for the Phase 2b 809/770 combo study and only for about half of the population.
Reasons justifying continued:
1. The interim results from the Phase 2b 809/770 combo study support this hypothesis because the stat. sign. SC improvement occurred only in the 809 phase and yet the FEV improved significantly, but only for about half the population. To me, this indicates that segment of the f508d population does not have a problem with sequencing and/or gating, or that the problem is minimal and can be overcome.
2. For a small subset of the double-deltas, there was an improvement when dosed solely on 770 (Kalydeco), indicating that subset in fact has protein on the cell surface. The improvement was not maintained, though, so I posit that the folding/sequencing defect prevents the protein from getting to the cell surface quick enough for potentiation. The initial improvement occurred because there was some protein there waiting for potentiation. Logically, this would seem to also explain why a longer (and possibly higher) dosing of 809 would result in such drastic improvement in Phase 2b–because now there is much more protein on the cell surface waiting to be potentiated, or for some, not needing to be potentiated.
1. Some Cfers on the 661 and 809 combo studies have indicated feeling significantly better on the 809/661 portion of the study, indicating they might not have an issue with gating.
1. The potential for success in the 661/809 combo study is much higher than originally thought because of the large portion of the f508d population retaining some gating function (i.e. having unlimited credit!)
2. While it is possible that some double deltas could benefit from 809 mono-therapy, because it is impossible to know which individuals have solely a folding defect (and/or a sequencing defect), combined therapy is the best approach. Also, even those individuals who benefit greatly from 809 or another corrector because they have gating function, the gating function might not be perfect, so a combo make sense.
3. There will be a need for the third med (what I call a sequencer) to address the portion of f508d Cfers whose protein also cannot fold in the proper sequence. This sequencer will benefit some double-deltas more than others (i.e., the 50%ish who did not have substantial FEV reduction will need this added med the most, while those with >10 will be least impacted, although they might not have perfect sequencing function.
Wow. That was a pretty technical post. I liked the club analogy.
Wasn't the purpose of the Phase II trails to measure a great improvement in SC? This was not accomplished. Vertex does not want the focus on the primary goal of the research which is where it was focused as a large improvement in SC would be in the right direction.
Highlighting the FEV ( I can breathe better ) improvement of ~ 5% seems weak to me. I would argue that a good steriod-based inhaler would increase the FEV more, but the test was not measured against all known forms of treatment ( antibiotics, inhalers, etc. ), it was measured against a placebo ( a control created to do absolutely nothing on purpose).
I am as hopeful as the next person in medical advancements and cures, but these tests are just not there yet. I think you can complicate the argument with deep-lingo science terms that only another geneticist would appreciate, but in the end, you need to create something that is orders of magnitude better than current treatment options. Vertex has not gotten there yet, maybe they will someday.
I can't buy into the idea that just because we learned how to create a telescope today, tomorrow we will all be traveling the galaxy. I can imagine it, I can hope for it, I can dream about it. But is a telescope really going to help me create a interstellar spaceship or will it just motivate me to want one.
Verity, I assume that the crux of the question you want to answer with your hypothesis I has to do with the diverse clinical response to VX-809.
Accumulation of abnormal proteins can bring about a major toxic burden for the cell and even compromise the organism's viability. Hence, adaptation and survival have forced evolution to create ‘quality control’ mechanisms, called [conformational checkpoints], that detect, monitor, and degrade such abnormally folded gene products.
Apparently, a majority of F508del mutant CFTR proteins is degraded even before reaching the cell membrane. This is a distinct property of 508 unlike that of 551. I would like to think that VX-809, sort of, mends the missing link at the phenylalanine at the 508th position, so that the checkpoint mechanisms find some of the repaired CFTR to have a close-to normal conformation ( = structure of protein folding). Then, some of VX-809 treated CFTR can survive degradation and would reach the cell membrane.
Now, the central question you are asking is that why clinical phenotype is good for some but mediocre for others. It has been established that 508 mutant is unstable at the body temperature even if it survives degradation (it is stable at 25C). My hypothesis here is that 809 mended 508 CFTR assumes diverse conformation because of remaining instability of the conformation, some will have intact channel activity and others less without the help of Kalydeco(the missing phenylalanine resides near the interface of cell membrane and the cell interior). This explains why VX-661 and 809 may have positive clinical phenotype without Kalydeco. The diverse conformation for the mended CFTR proteins even with the presence of Kalydeco would make the overall phenotypes to be diverse.
The above is my explanation as to why the same genotype does not lead to the same phenotype even with the help of the drugs.
Thank you so much for your insights and for so kindly taking the time to explain this all to me–I really wouldn’t be able to understand much of this without your kind teaching.
You actually made me think more about “what” was the underlying crux of this hypothesis, and I think there are 3 things I’m trying to get at: 1) As you note, why are there diverse clinical responses?; but also 2) when we say “why” were there diverse clinical responses, not such the mechanism, but the underlying reason for the mechanism that caused the diverse responses and then what can be done to improve those responses; and 3) why were some of those responses so strong–stronger than in vitro suggested and Vertex expected.
Initially, I am struggling with a few things to understand your explanation. I completely understand what you are saying about the adaptation and survival/confromational checkpoint. (That helps so much b/c I didn’t quite get the misfolding thing but now I do!) But I’m struggling though with understanding how the channel activity relates this. Is the channel activity a second “checkpoint” to prevent defective protein from moving once at the cell surface, similar to the degradation response that prevents the misfolded protein from moving to the cell surface in the first instance? So with 551, the protein isn’t misfolded so it makes it past the first “checkpoint” but it is still defective, so it is not allowed to “channel?” I guess I’m wondering if there is a second thing that I’m missing here related to the channeling? Knowing this will help me understand your explanation and also to think through more why some response to 809 alone but not Kalydeco.
But if I am understanding what you wrote, then:
1) The reason for the diverse clinical response is that everyone’s body has a different survival checkpoint, i.e., “close enough” to “conformation” for one person is not good enough for another? (To use the “clubbing analogy”–different bouncers have different standards for “hot chicks?”) And relatedly, some of those “close enough” to “conformation” are seen by some Cfers’ bodies as “close enough” to channel freely; and thus no need for Kalydeco. Also, even some of those “close enough” proteins will later be rendered “rejects” because of further degradation, which also can vary be person. Is this right???
2) So then the second question is “why” is “close enough” okay for some and not others? I don’t understand evolutionary science much, but is merely part of that science that different bodies have higher/lower standards for “defective” genes, etc.? But I also wonder if there are other things that could affect. For instance, you mention body temp as causing the degrading: Could having a higher basal body temp/lower temp (hypothyroidism or is it hyper?) matter. My basal temp is 1 degree lower than the “norm” (thyroid issues)–so could things like that affect degradation? If that is the case, then maybe there is a way to “improve” 809's impact by either making more proteins look close enough (or prevent them from degrading) or altering how the body sees “close enough.” (Again, clubbing analogy–if you put a lot of make-up on, some might make it past the first “hot chick” checkpoint, or give enough beers to the bouncer.)
3) Finally, why were some of the results much better than I think Vertex expected–10% is basically Kalydeco and I believe Vertex said in vitro Kalydeco had fixed 50% of proteins while 809 only 30%. I wonder if it is either/both because: a) in vitro they were looking for “perfect” proteins, while the body only needed “close enough” and b) improved transcomplementation between the now-improved proteins.